Perceptions and experiences of Australian speech-language pathologists who use the Lidcombe Program with children who stutter

Abstract

Purpose: Speech-language pathologists (SLPs) have varying levels of training and experience working with children who stutter (CWS). They often work in contexts and populations that require clinical management to take them beyond the borders of translating clinical trial protocols and evidence-based practice (EBP). This study sought to investigate the clinical experiences of Australian SLPs working with CWS, including their Lidcombe Program confidence and practices. Method: A cross-sectional online survey was completed by 215 Australian SLPs during 2017–2020. They were recruited prior to attending one of 14 workshops hosted by Speech Pathology Australia (SPA). Result: Participants worked in a range of contexts and were from all states and territories. The majority assessed and/or treated up to five CWS annually. Six predictor variables were considered for self-reported clinical confidence. The quantitative analysis identified that an SLP’s previous training, reading the SPA stuttering management clinical guideline, and the annual number of CWS treated were found to significantly impact self-reported clinical confidence. Themes included: (1) client, family, service, and context factors that influence clinical decision-making; (2) SLP areas of clinical strength and aspects of paediatric stuttering management requiring further development; and (3) factors that impact Lidcombe Program success and modifications. Conclusion: This study has found that Australian SLPs working with CWS identify a range of important factors that impact their practice.

Keywords:

• clinical practice
• Lidcombe Program
• children who stutter
• stuttering
• confidence

Introduction

Stuttering is a complex communication disorder that can negatively impact a child’s life (Brignell et al., 2021; Briley, 2018; Nye et al., 2013; Sugathan & Maruthy, 2021). Some parents of preschool children report they appear impacted by their stuttering (Boey et al., 2009; Langevin et al., 2010). School-age children who stutter (CWS) are reportedly at risk of being teased or bullied, having negative communication attitudes, and developing social anxiety (Guttormsen et al., 2015; Rocha et al., 2019). Stuttering onset occurs between the age of 2–4 years and has a cumulative incidence of approximately 11% at 4 years of age (Reilly et al., 2013). It is not possible to accurately predict if CWS will experience natural recovery (Baxter et al., 2016; Briley, 2018; Byrd & Donaher, 2018; Lowe et al., 2021; Nonis et al., 2022; Reilly et al., 2013; Sugathan & Maruthy, 2021). Consequently, treatment is recommended for CWS (Baxter et al., 2016; Brignell et al., 2021; Laiho et al., 2022). The published literature focuses on intervention, with little reported on assessment practices. Brundage et al. (2021) developed consensus guidelines for stuttering, which include recommendations for paediatric populations. These include parent-reported: (a) case history focusing on stuttering-related information; (b) speech, language, and temperament development; (c) fluency and stuttering behaviours; (d) reactions to stuttering by the child and others in their environment; and (e) observations of adverse impact.

Clinical translation of evidence based treatment for CWS

While paediatric stuttering treatment has been widely investigated, recent systematic reviews have found that study quality varies (Brignell et al., 2021; Hofslundsengen et al., 2022; Nye et al., 2013; Laiho et al., 2022). Most research has focused on the preschool population, rather than school-aged population, and direct behavioural treatments, which have reduction in stuttering as the primary outcome, compared to indirect approaches that target environmental triggers rather than stuttering behaviours (Donaghy & Smith, 2016). To date, the literature reports 19 treatments for CWS (Hofslundsengen et al., 2022), with the bulk of the research investigating the Lidcombe Program (Brignell et al., 2021; Hofslundsengen et al., 2022; Nye et al., 2013). The Lidcombe Program is a two stage direct behavioural treatment. During Stage 1, the speech-language pathologist (SLP) teaches the parent to complete daily treatment and monitor severity ratings until stuttering reduces consistently to low levels (Onslow et al., 2020). The child can then enter Stage 2, whereby treatment is gradually withdrawn across less frequent sessions, with the aim to maintain nil-to-low levels of stuttering.

Postgraduate training in the Lidcombe Program is available within Australia and internationally under the auspice of the Australian Stuttering Research Centre (O’Brian et al., 2013). This typically involves a 2 day workshop delivered by Lidcombe Program consortium trainers. While training is recommended, it is not mandatory as the Lidcombe Program training manual is available freely online. The significance of being Lidcombe Program trained has been investigated in two community implementation studies (O’Brian et al., 2013; O’Brian et al., 2023). In the initial 2013 study of 31 Australian SLPs, 45% reported that they had attended consortium training. This training was found to be a significant predictor of client stuttering reduction at 9 months post-commencement of treatment. The mean stuttering frequency for clients of trained SLPs was 1.1% syllables stuttered (SS) compared to 2.4%SS for untrained SLPs. The subsequent 2023 study of 51 Australian SLPs and 15 UK SLPs did not establish a difference in client outcomes based on SLP Lidcombe Program training.

While good quality research is available to guide SLPs working with CWS, direct translation of clinical trial protocols to typical clinical settings can be challenging. Some of the reasons for this include: (a) the heterogeneity of clinical caseloads and families when compared to the homogeneity of research participant groups (Brignell et al., 2021; Nye et al., 2013); (b) the contrast in timing and resource constraints of community SLP services compared to research trials, which can provide the required number of sessions at the specified frequency and duration (Lowe et al., 2021); (c) the skills, knowledge, confidence, and education of the SLP (Briley, 2018; Byrd et al., 2020; Coalson et al., 2016; Lowe et al., 2021); and (d) the lack of direct access to peer-reviewed journal articles for many community SLPs (Lowe et al., 2021). Hofslundsengen et al. (2022) posit that compliance with treatment protocols may be a contributing factor in the reported treatment outcomes in a study, rather then the discrete treatment itself. Further issues reported in intervention studies relate to researchers not considering the constraints on community service models when developing clinical trial protocols, resulting in treatment programs that are difficult to translate into practice (Brignell et al., 2021; Langevin & Kully, 2003; Lowe et al., 2021).

Speech-language pathologists working with children who stutter

It is well established that many paediatric SLPs have large caseloads requiring a broad and general skill set (Chmela & Johnson, 2018; Nippold & Packman, 2012; Oyler & Chmela, 2003). Several survey studies were completed during the past decade to explore the knowledge, skills, education, and confidence of SLPs who work with CWS (Beita-Ell & Boyle, 2020; Briley, 2018; Byrd et al., 2020; Coalson et al., 2016; Rousseau et al., 2002; Tellis et al., 2008). Collectively, these studies pool data from more than 1600 SLPs, of which 70% worked in US schools (Beita-Ell & Boyle, 2020; Briley, 2018; Byrd et al., 2020; Coalson et al., 2016; Tellis et al., 2008). Key themes that emerged from these studies included SLPs’ lack of confidence, knowledge, training, and clinical skills in working with CWS (Briley, 2018; Byrd et al., 2020; Coalson et al., 2016; Rousseau et al., 2002; Tellis et al., 2008), and that SLPs have limited and varying clinical exposure to CWS (Beita-Ell & Boyle, 2020; Briley, 2018; Tellis et al., 2008).

The perspectives of specialist paediatric stuttering SLPs have also been investigated (Boyle et al., 2021; Byrd & Donaher, 2018; Lowe et al., 2021; Van Eerdenbrugh et al., 2018). Like the generalist SLPs, there is broad recognition of the barriers that exist in translating clinical trial evidence into typical clinical settings and populations—an understanding that treatment needs to be individualised for clients and their families, and an acknowledgement of the need for ongoing professional development.

Present study

To date there appears to be a range of factors that influence the clinical translation of research for SLPs working with CWS. It is therefore necessary to both understand the practices, perceptions, and experiences of SLPs working with CWS in Australia, and compare these to what has been established in other jurisdictions (Beita-Ell & Boyle, 2020; Briley, 2018; Byrd et al., 2020; Chmela & Johnson, 2018).

Therefore, the aims of this study are to:

• 1. Investigate the assessment and treatment practices of Australian SLPs who work with CWS.

• 2. Investigate the role of the following six variables on SLP confidence and practices with assessment, education, and delivery of the Lidcombe Program: SLP clinical context, completion of Lidcombe Program training, recency of Lidcombe Program training, reading of the Speech Pathology Australia clinical practice guidelines for stuttering management, estimated total number of CWS assessed annually, and estimated total number of CWS treated annually.

• 3. Identify SLP reported barriers and facilitators to stuttering treatment success.

• 4. Establish SLP identified knowledge and skills gaps related to assessment, education, treatment, and delivery of the Lidcombe Program to inform professional development workshop learning objectives.

Method

The methodology for this study is replicated from Serry and Levickis (2021) who completed a cross-sectional survey of SLPs prior to their attendance at one of 11 national literacy workshops hosted by Speech Pathology Australia in 2017. They reported a response rate of 51%. Ethics approval for this study was obtained through La Trobe University Human Research Ethics Committee (approval number HEC20294).

Participants and recruitment procedure

Participants were SLPs who attended one of 14 paediatric fluency professional development workshops presented by either the first author alone or both authors together, between October 2017 and June 2020. The half-day or full-day workshops were presented throughout Australia and provided a broad overview of paediatric fluency practice. Topics included the latest research findings, a summary of the Speech Pathology Australia clinical practice guideline for stuttering management (Speech Pathology Australia, 2017), and assessment, measurement, and treatment principles and practices. SLPs self-registered for the workshops. There were no exclusion criteria for workshop attendance.

Survey development and pilot testing

The survey was developed for the purposes of this research. Thus, similar to previous studies (Beita-Ell & Boyle, 2020; Boyle et al., 2021; Briley, 2018; Byrd et al., 2020; Coalson et al., 2016; Rousseau et al., 2002; Tellis et al., 2008), a non-standardised survey was used to ensure questions were suitable for the target population, treatment practices, and contexts. The survey included both closed-ended questions presented using a Likert scale, and open-ended questions where a more detailed explanation was required. An initial preworkshop survey was trialled by the first author with one metropolitan workshop cohort (n = 7) and one regional workshop cohort (n = 16) in 2016. Feedback from these cohorts was used to determine the validity of the survey and ensure the clarity of the survey questions. The second survey iteration was reviewed independently by two research academics in different states. Their feedback did not result in further changes to the survey.

Final survey distribution and response rate

All workshop registrants (n = 325) were invited to complete the survey. They were emailed a link to the survey 1 week prior to attending and the survey closed the day before the workshop (Supplementary File 1). The email was sent from the workshop hosts who were independent of this research study. The survey was anonymous and completed using the Qualtrics online platform, meaning the researchers were unaware of which attendees at the workshops had completed the survey. This was to reduce the risk of social bias informing both participation and the nature of the survey responses.

In total, 215 SLPs completed the survey and consented to participating in this study. This represents a 66% response rate, which is a high response rate for an online survey given the usual response rate for comparable methodologies. Typically, where a single invitation without repeated follow-up is used, response rates are reported to vary between approximately 16% and 34% (Brtnikova et al., 2018; Kidd et al., 2021; Poynton et al., 2019). Our response rate is also comparable to the 75% response rate reported in the Rousseau and colleagues’ (2002) survey of SLPs delivering the Lidcombe Program in Australia, and the 51% response rate reported by Serry and Levickis (2021) who completed a cross-sectional survey for participants of 11 Speech Pathology Australia workshops. Finally, given our inclusion of open-ended questions in the survey, our large sample size meets the recommendation of Braun et al. (2021) of a sample of at least 50 participants to allow thematic analysis of this data.

Demographic data

Participants were based in each Australian state and territory, with the majority from Victoria (33%), Queensland (20%), and Tasmania (19%). The participants’ work contexts spanned health, education, disability, early intervention, private practice, not for profit, and university clinics. However, most participants worked in private practice (46%), a health context (24%), or an education setting (19%). There was a near perfect split between those participants who had participated in formal Lidcombe Program training (51%) and those who had not (49%). For those who had been trained, this had occurred anywhere from within the last 2 years through to more than 10 years prior, with an even spread of participants. Almost three-quarters of the participants (71%) had read the Speech Pathology Australia stuttering clinical guidelines (Speech Pathology Australia, 2017). The vast majority of participants only engaged in stuttering assessments and treatment as a component of their overall workload. Most completed fewer than five assessments per year (66%) and similarly, most treated fewer than five children per year (70%). Approximately 15% of participants completed more than 10 assessments per year and a similar proportion treated more than 10 CWS per year.

Data analysis

Quantitative data

There were six potential explanatory variables (independent variables) and six response variables of interest (dependent variables) in the survey that were analysed. The explanatory variables were: (a) the participants’ work (clinical) context, (b) whether they had received Lidcombe Program training, (c) when the training had occurred, (d) whether the participants had read the Speech Pathology Australia stuttering clinical guideline, (e) how many stuttering assessments they completed per year, and (f) how many clients they provided treatment for per year. Five of the six response variables of interest related to the self-reported level of SLP clinical confidence. These included: (a) confidence to deliver education to clients and families regarding preschool stuttering, (b) confidence to deliver education to clients and families regarding school-age stuttering, (c) confidence to conduct a paediatric stuttering assessment, (d) confidence delivering Stage 1 of the Lidcombe Program, and (e) confidence delivering Stage 2 of the Lidcombe Program. The final variable of interest was the self-reported success rate of clients completing Stage 1 of the Lidcombe Program.

Since the response variables are rating scales and thus ordinal, the ordinal logistic model, which is also known as the proportional odds model, was used (Agresti, 2013). Before model fitting commenced, each categorical predictor was cross tabulated with each response variable to check for empty cells that may mean model parameters are not estimable or at least may be very unreliable (Agresti, 2013). A cross tabulation revealed empty cells for “Work context” for all six response variables. To counter this problem, the categories “Student”, “Not for profit”, “Disability”, and “Other-not specified” were combined into an “Other” category. A strong indication of collinearity between “Assessments per year” and “Treatment clients per year” was detected, meaning that they essentially convey the same predictive information. A simple solution to this problem was to omit one of the variables from the model and in this case “Assessments per year” was omitted. Another variable pair to exhibit collinearity was the nominal predictor “Lidcombe Program training” and the ordinal predictor “Years since Lidcombe Program training”. The latter was omitted as it conveyed no significant additional predictive information over that of “Lidcombe Program training”. One final issue was the multivariate nature of the responses. Since each respondent answered all six questions, significant correlations existed between them. As such, the multivariate method was required to avoid bias in parameter estimates.

Qualitative data

Data analysis

Open-ended responses were analysed using Graneheim and Lundman’s (2004) procedure for qualitative text responses in health research surveys. This method of qualitative content analysis takes a structured approach where survey responses are first broken into meaning units (text related to each other through content and context), condensed where the description retains the key text and quality, and then labelled (coded). Next, the content is interpreted for meaning, grouped into catgories (subthemes), and finally broader themes (meaningful threads) are established.

The analysis began with the first author initially reading the survey responses several times to gain an overall understanding of the data. Using the qualitative coding software program NVivo, the text was divided into meaning units that were condensed and labelled with a code. Tentative generic categories were generated and named to classify and organise the open codes. For example, open codes included: clinic sessions and home treatment is often shorter and delivered less often, a non-parent agent is sometimes used to administer treatment, some CWS attend sessions without parents present, and the program is sometimes supplemented by non-Lidcombe Program strategies. A generic category labelled “the Lidcombe Program is modified (or adapted) to varying degrees” was then generated to group these codes. While the survey inherently included specific questions about the SLPs’ practices and perceptions, the categories themselves were generated from the data using the process described. A frequency count was completed by the first author to identify the non-Lidcombe Program treatments SLPs reported using (Kalaian, 2008).

Rigour criteria

Thoroughness and rigour were achieved through several strategies. Dependability was addressed by keeping an audit trail for transparency of methods and data analysis. In addition, coding accuracy was measured by having an SLP independent of the project complete the aforementioned data analysis procedure on 13% of the survey data. Once completed, themes were reviewed with the first author and agreed upon by both collaboratively (McAllister & Lyons, 2019). Credibility was established, as previously reported, through pilot testing the survey with 23 SLPs and two research academics who work with children who stutter. Additionally, both authors are experienced researchers and clinicians in the field who met regularly during the data analysis to establish agreement about the development of themes, and to ensure that no relevant data were inadvertently excluded. Credibility was also addressed through the use of verbatim quotes from the surveys in the results section.

Result

Quantitative data

Predictors of clinical confidence and performance

Preliminary analyses included all six predictors, and each was assessed for multivariate significance using a likelihood ratio chi-squared test. Since the p-values for “Work context” and “Years since Lidcombe Program training” were well above 0.10, they were removed from all models. The remaining predictors (“Lidcombe Program training”, “Reading the Speech Pathology Australia clinical guideline”, and “Number of treatment clients per year”) were all statistically significant and were retained in all models.

The predictor coefficients were calculated to quantify the effect of a categorical predictor relative to its base level on the log-odds scale. It is usually easier to interpret odds ratios, which are the exponentiation of the coefficients, and these are presented in Supplementary File 2 along with the associated p-values. Odds ratios provide a measure of association. Using the data from the first row, a participant trained to deliver the Lidcombe Program is 5.6 times more likely to have higher self-reported confidence delivering education to clients and families about preschool stuttering than a participant without Lidcombe Program training. Since all Lidcombe Program training p-values are less than 0.05, similar statements about the effect of training apply for all responses.

The odds ratios and p-values for the predictor “Read guide” show that a participant who has read the Speech Pathology Australia clinical guideline had a statistically significant probability of a higher self-reported confidence delivering education to clients and families about preschool stuttering than a participant who has not read the guideline. However, the p-values for all other responses are greater than 0.05, indicating that reading the Speech Pathology Australia clinical guideline has no statistically significant probability of impacting the participants’ self-reported confidence providing school-age stuttering education and Stage 1 and 2 of the Lidcombe Program, or self-reported Lidcombe Program Stage 1 success rate.

The odds ratios and p-values for the predictor of number of clients treated per year show that a participant who treats six to 10 clients per year has a statistically significant probability of having higher confidence for preschool stuttering education, and confidence delivering Stage 1 of the Lidcombe Program, than a participant who treats zero to five clients per year. Treating six to 10 clients has no statistically significant impact on the probability of a participant’s confidence delivering school-age stuttering education or self-reported Lidcombe Program Stage 1 success rate relative to participants who treat zero to five clients. Lidcombe Program Stage 1 success refers to the SLP’s recount of their client completing Stage 1 as detailed in the Lidcombe Program treatment guide (Onslow et al., 2020).

The odds ratios and p-values show that treating 11 to 20 clients per year impacts on preschool and school-age education confidence but not on Lidcombe Program Stage 1 or Stage 2 confidence, or self-reported Lidcombe Program success rate relative to those with zero to five clients. Participants who treat more than 20 clients per year have a statistically significant probability of having higher confidence providing education, and higher confidence in delivering Stage 1 and 2 of the LP, than participants who treat zero to five clients for both preschool and school-age groups. However, there is no statistically significant difference in the probability of self-reported Lidcombe Program treatment success rate (p = .06).

In addition to odds ratios, individual cell probabilities were calculated for participant self-reported Lidcombe Program Stage 1 success rate. The other included variables are reading the Speech Pathology Australia clinical guideline, the number of clients treated per year, and the self-reported Lidcombe Program success rate. As an example, we can report that a participant who has (a) read the Speech Pathology Australia clinical guideline, (b) treats between zero and five clients annually, and (c) reports a greater than 75% success rate is 3.6 times more likely to have received Lidcombe Program training than not. The ratio of probabilities for those with and without Lidcombe Program training generally increases as the response level increases. In other words, those having completed Lidcombe Program training exhibit statistically significant higher probability of self-reported success rates than those without Lidcombe Program training. The impact of reading the Speech Pathology Australia guide and the number of fluency clients treated are significant but are less noticeable.

Treatment approaches used other than the Lidcombe Program

SLPs were asked to list any treatment approaches in addition to the Lidcombe Program that they use with CWS. One hundred and nine SLPs reported alternative treatments used for preschool children and 105 SLPs reported alternative treatments used for school-age children. There was no limit to the number of treatments SLPs could list for each age group. Twenty per cent of SLPs reported using treatments other than the Lidcombe Program for preschool CWS and 40% use non-Lidcombe Program treatments with school-age CWS. Use of these programs may or may not be in addition to the Lidcombe Program. Table I details the treatments used and frequency.

Table I. Treatment approaches and frequency used by Australian speech-language pathologists (SLPs).

	Preschool (n = 109 SLPs)	School-age (n = 105 SLPs)
Direct	Lidcombe Program only (66) Speech restructuring strategies (9) Syllable timed speech (8) Modified Lidcombe Program (3) Extended length of utterance (2)	Lidcombe Program only (20) Speech restructuring strategies (51) Syllable timed speech (26) Timeout (14) Self-imposed (13) Therapist imposed (1) Extended Length of Utterance (2) Other (4) Stop and think (1) Deliberate stuttering (1) Airflow management (1) Positive reinforcement (1)	Hybrid (5) Lidcombe Program and speech structuring strategies (2) Lidcombe Program and “stop and start” and “take a breath” (1) Combination of Lidcombe Program and extended length of utterance (1) Westmead “with adaption for naturalness” (1)
Indirect	Demands and Capacities Model (DCM) General DCM strategies (13) Palin Parent-Child Interaction (5) RESTART-DCM (3)	Communication attitudes (2) Demands and Capacities Model General DCM strategies (2) Palin Parent-Child Interaction (1) RESTART-DCM (1)
	Other Ensuring speech and language skills appropriate (1)
Combined indirect  + direct	Response contingent stimulation (4) Other Easy Does It (1) Yaruss Family Focused therapy (1) Home program (1) Ensuring speech and language skills appropriate (1)	Other Easy Does It (2) Yaruss Family Focused therapy (1) Home program (1)

Note. Direct refers to those treatments where the primary outcome is reducing stuttering; Indirect refers to those treatments that target environmental triggers rather than stuttering behaviours (Donaghy & Smith, 2016). Combined Indirect + Direct refers to those treatments that include elements of direct and indirect approaches.

Qualitative data

Three overarching themes emerged from the data. Themes 2 and 3 also contained subthemes.

Theme 1: Speech pathologists identify client, family, service, and context factors that influence clinical decision making in paediatric fluency

The participants identified a myriad of factors that impact their clinical decision making. First, a client’s age and presence of concomittant disorders informed SLPs’ decisions relating to the timing of treatment and the type of treatment considered. Specifically, the client’s age related to whether the children were preschool or school-age, defined as 0–6 and 6–12 years in the survey, and concomittant disorders referred largely to other communication disorders, namely speech sound disorders or expressive, receptive, or mixed language disorder. A number of SLPs also identified a diagnosis of autism spectrum disorder (ASD), attention-deficit/yperactivity isorder (ADHD), intellectual disability, and anxiety as co-morbid presentations that are considered as part of their clinical decision making. For example, “working with co-morbidity of language and fluency difficulties” and “treatment of stuttering and coexisting disorders such as ASD, speech and language”.

Second, SLPs identified parent and family capacity as a key factor, with the home environment and family schedule impacting engagement in the treatment process. In many cases SLPs identified that parents and families have competing demands that impact the time or resources they have available to implement treatment recommendations. SLPs also acknowledged that treatment and rating stuttering was difficult for some parents to learn or consistently demonstrate. Several SLPs touched on “culture” and “socio-economic status” as being variables that can affect an episode of care. It was unclear how exactly, beyond the quotes provided, these factors impacted treatment. For example, “cultural and family dynamic factors obviously affect the ways in which therapy is applied - I try to tread a line where I respect their modes and preferences while keeping a strong eye on the literature”, “socio economic status and the challenges of sticking with the program with other competing issues”, and “parents understanding the need for regular, and specific practise in the home setting”.

Third, services may offer treatment in treatment blocks, groups, a limited number of sessions, or provide a review service only: “we are running groups”, small school setting with limited visits over time”, and “deliver in blocks - 6-8 weeks”.

Finally, SLPs may provide treatment by telepractice, home visits, or in education settings, or decide to refer children to an alternate or specialist stuttering service: “offering sessions over Skype for rural school”, try to assess and treat clients in realistic settings i.e. classroom rather than clinic”, and when is it ok to say - our service won’t help you, need to look into private tx [treatment] (most families can’t afford this option)”.

Theme 2: SLPs identified areas of strength and areas requiring further development

As fluency clinicians, the participants identified a number of clinical strengths. These included: (a) requesting and reviewing parent ratings in each session; (b) modelling treatment; (c) establishing rapport; d) encouraging, training, and giving feedback to parents; (e) providing education about stuttering and treatment requirements; (f) assessing stuttering and using stuttering measures; and (g) adhering to Lidcombe Program treatment guidelines and session structure. The participants also identified aspects of the assessment and treatment process where they would like further education or skill development. These included assessment practices, timing and choice of early treatment, and general management of fluency caseloads. Figure 1 shows each further education subtheme and the condensed meaning units.

Figure 1. Theme 2 subthemes and condensed meaning units (as per Graneheim and Lundman’s, 2004, procedure for qualitative text responses in health research surveys described on page 12). ¹School-age children who stutter.

Figure 2. Theme 3 subthemes and condensed meaning units (as per Graneheim and Lundman’s, 2004, procedure for qualitative text responses in health research surveys described on page 12). ¹Children who stutter.

Theme 3: Factors that impact Lidcombe Program success and modifications

The participants attributed Lidcombe Program success to the efficacy of the program on stuttering, family factors, and service delivery. That is, some SLPs indicated the Lidcombe Program works purely as a result of its design, suggesting there is an intrinsic or even unknown reason for its success. A vast majority of SLPs simply identified not being able to translate the Lidcombe Program as it is prescribed. The two main reasons that emerged related to (a) their workplaces not being funded for, or allowing, ongoing weekly 45–60 minute sessions or (b) that parents were non-compliant with completing the prescribed daily severity ratings and 10–15 minutes treatment until Stage 1 criteria were met (nil-to-low level stuttering observed both beyond session and within session). In contrast, SLPs that reported success with the program referred to parents completing the recommended treatment and ratings. For example, my clinical environment [is] well suited to delivery of program according to guidelines”, “have not had the opportunity to implement the Lidcombe Program due to restraints of working in a school based setting with minimal parent involvement” and “limitation of number of weekly clinic sessions that can be offered in CH [Community Health Service] often means that ‘Stage 2' is entered not when a client is ready, but when weekly sessions can no longer be offered.”

The participants acknowledged that they modify the program to varying degrees, including sometimes adapting or omitting Stage 1 session elements as stated in the Lidcombe Program treatment guide (Onslow et al., 2020). Stage 1 session elements include eliciting a child speech sample and co-rating with parents; a discussion of the previous week’s daily severity ratings and treatment; parent demonstration of treatment; SLP feedback and coaching to parents regarding the demonstration; SLPs trialling and demonstrating planned treatment changes for the prospective week; and planning discussion for the week ahead including treatment sessions, activities, or particular foci. SLP omissions are exemplified in the following quotes: sometimes I skip the parent rating of stuttering”, “I often have to skip demonstrating to parents, because they’re not there”, “I can use principles of a Lidcombe Program if family circumstances do not allow for sessions outlined in Lidcombe Program”, and “occasionally use Mr Bumpy and Mr Smooth as this has been useful prior to gaining knowledge and informal training of the Lidcombe Program” (Figure 2).

Finally, the participants identified 11 aspects of the Lidcombe Program treatment process that they would like to review in a professional learning context. These were: (a) teaching, coaching, engaging, and educating parents; (b) managing parent compliance, motivation, and capacity; (c) when to adapt, cease, or modify; (d) time, duration of sessions, dose, and treatment; (e) service delivery options; (f) practical tips and real-life examples; (g) how to manage slow response, plateau, or regression; (h) how to be confident and administer the program effectively; (i) specific details and recommendations about verbal contingencies; (j) Stage 2 criteria, session structure, and treatment schedule; and (k) how long to monitor for and when to discharge.

Discussion

It is well established that the provision of effective early treatment is crucial to reducing the potential long-term negative effects of stuttering (Baxter et al., 2016; Brignell et al., 2021; Nye et al., 2013). Yet many SLP studies and expert commentaries report a myriad of issues impacting the translation of clinical trial treatment programs (Beita-Ell & Boyle, 2020; Briley, 2018; Byrd & Donaher, 2018; Byrd et al., 2020; Coalson et al., 2016; Lowe et al., 2021; Rousseau et al., 2002; Tellis et al., 2008). It is unclear, however, whether issues reported internationally relate to the contemporary practice of Australian SLPs. Therefore, this study explored Australian SLP assessment and treatment practices; possible predictor variables of self-reported confidence relating to stuttering assessment, education, and treatment; SLP identified barriers and facilitators for treatment success; and self-identified learning needs.

The current demographic of Australian SLPs who work with CWS

Demographic data reported from the 215 SLPs who participated in this cross-sectional survey study provide significant insight into current paediatric practice with CWS across Australia. When comparing results with the two published Australian SLP Lidcombe Program studies (Rousseau et al., 2002, and O’Brian et al., 2013), there are several similarities and differences. First, despite the widespread use of the Lidcombe Program, in each of the three studies only approximately half of SLPs had received formal training. Despite the proliferation of Lidcombe Program research and the widespread dissementation of information about the program, it is noteworthy that there is little difference in training rates between participants in our study and those in Rousseau and colleagues’ research that was published nearly 20 years ago. This lack of SLP Lidcombe Program training is likely to impact clinical outcomes of CWS given O’Brian and colleagues (2013) found a positive correlation between SLP training and client stuttering reduction 9 months after commencing treatment. Similarly, participants in our study who self-reported a Lidcombe Program Stage 1 success rate of at least 75% were much more likely to have undertaken training than not. There are also similarities with the Australian SLPs in the O’Brian et al. (2023) translation study, where 39% of the SLP participants reported being Lidcombe Program consortium trained. These similarities include the SLPs’ work context and the modifications to the Lidcombe Program typically made. The SLPs in the four studies reported common adaptations to the Lidcombe Program including non-weekly treatment sessions, shorter session duration, omission of parent treatment demonstrations, and absence of measurement. This similarity could partly be due to the similar proportions of participants who had received Lidcombe Program training in each study, the common translation issues that were reported across all studies, or because the translation issues in Australian speech pathology remain unchanged. O’Brian et al. (2023) refer to these issues as “therapist drift”.

Further comparisons can be made between this study and the Rousseau et al. (2002) study. First, in both studies, the majority of SLPs had limited exposure to CWS and small caseloads. This study found that approximately two-thirds of participants assessed and treated between 0–5 CWS annually, compared to Rousseau et al. (2002) who found 50% of their SLP participants saw between 1–10 CWS in the 3 years preceding the study. This finding reiterates that Australian SLPs who treat CWS continue to largely be generalist paediatric SLPs. A similar finding has been reported in several US studies. For example, Beita-Ell and Boyle (2020) reported a mean annual caseload of 1.67 CWS for school-based SLPs; Briley (2018) reported a mean caseload of 2.2 CWS annually for school-based SLPs; and Tellis et al. (2008) reported an annual mean of two CWS, with a mean career total of 12. In contrast, Boyle et al. (2021) reported a mean yearly caseload of 43.34 CWS and a mean career total of 550.74 CWS for board certified, specialist fluency SLPs.

A key demographic difference between this study and the participants in the Rousseau et al. (2002) study is the SLPs’ workplaces. This is significant not only to establish SLPs’ workforce trends, but also to better understand the service contexts in which CWS currently have access to treatment and the implications this may have on clinical translation. Almost half of the SLPs in this study reported working in private practice (46%) compared to 27% in Rousseau et al. The proportion of participants working in health services was 46% in Rousseau et al. and only 24% in the current study. This is considerable given private practice in Australia generally relies on a full fee recovery model, while health services are free or heavily subsidised. This shift is likely to have had an impact on some of the significant barriers to clinical translation that are reported in this study, namely reducing the frequency or length of sessions.

Predictor variables for SLP self-reported confidence

SLP self-reported confidence in providing education, assessment, and Stage 1 and 2 of the Lidcombe Program was higher for participants who had completed Lidcombe Program training. Given the Lidcombe Program training is the main postgraduate training for stuttering available in Australia, it could be that attendance and engaging in any form of professional development generally improves confidence, rather than it being Lidcombe Program specific. If so, this supports professional commentaries relating to SLPs needing more access to professional development in stuttering (Byrd & Donaher, 2018; Lowe et al., 2021).

Assessment and treatment practices of Australian SLPs

Overall, SLPs reported being more confident in assessment and education than treatment for both preschool and school-age CWS. This is consistent with Briley (2018) who found 56% of school-based SLPs were extremely confident in screening CWS compared to only 14% who were extremely confident in providing treatment. Yet, many still identified assessment practices and stuttering measurement as areas for further learning. This fits with Byrd et al. (2020) who reported SLPs have a lower comfort level with diagnosing stuttering compared to other development communication disorders.

While this study confirmed the Lidcombe Program is the predominant treatment used for CWS who stutter in Australia, it also established that use, delivery, and outcomes differ when delivered in the community compared to clinical trial reports. This is evidenced in the varying level of self-reported confidence and success in completing Stage 1 and 2 of the program, and myriad of ways the program is modified (captured by Theme 3). While some modifications fit within the Lidcombe Program recommendations of individualising for the child and family, a significant majority deviate greatly. Such deviations include clinical sessions being shorter than the specified 45–60 minutes and omitting one or several of the prescribed Stage 1 session elements, namely severity ratings, SLP or parent treatment demonstrations, and parent coaching. Further, SLPs reported providing sessions less often than the prescribed weekly schedule, using a non-parent agent (e.g. a student aide at school etc.) to deliver treatment, or delivering treatment to the child in the absence of a parent.

The results of this study also indicate that Australian SLPs draw on a range of indirect and direct stuttering treatments. While the majority report using the Lidcombe Program for both preschool and school-age children, 20% report using a range of other treatments for preschool CWS and 40% for school-age children who stutter.

Clinical translation: Barriers and facilitators

Analysis of the open-ended survey questions revealed that the SLPs perceive Lidcombe Program success to be attributed to many factors. These factors can be largely categorised as family factors or service delivery factors. The SLPs highlighted encountering challenges associated with parental engagement, leading to problems with treatment compliance. This impacted both attendance at the clinic sessions and completion of daily home practice with the child. The SLPs also noted that the effectiveness of the Lidcombe Program can be affected by the capacity of parents to deliver the treatment accurately. This included reporting that some parents find the process of appropriately structuring the daily home practice sessions to meet the requirements of the program to be difficult. These findings replicate the parent-related variables reported by Rousseau et al. (2002) and substantiate the themes relating to parents in the study by Van Eerdenbrugh et al. (2018), which reported Lidcombe Program challenges identified in the literature and by expert Lidcombe Program SLPs. These parent-related challenges were also broadly identified by parents themselves in a recent systematic review of parent perceptions of stuttering (Nonis et al., 2022). It is important to highlight that parent-related variables, such as compliance with treatment requirements and accurate delivery of treatment, should be addressed by the treating SLP. A clear rationale presented to parents describing the link between the anticipated efficacy and efficiency of treatment (as outlined in the published research) and adhering to the treatment protocol is required. Additionally, the use of a checklist, such as the parent administration of the Lidcombe Program checklist presented by Swift et al. (2012) or the problem solving considerations highlighted by Van Eerdenbrugh et al. (2018), is recommended.

The SLPs also identified service delivery barriers impacting their ability to deliver treatment. Providing treatment to children transitioning from preschool to primary school is particularly challenging. The participants highlighted that eligibility for services often ceases upon a child commencing school. Additionally, where a child is eligible, their ability to engage in treatment is affected by time demands associated with schoolwork and social activities. For school based SLPs, these factors are intensified by trying to manage large caseloads and scheduling treatment sessions around school holidays. Such issues are consistent with the results of SLP survey studies reported in the introduction of this paper. This is also compounded by the lack of treatment evidence for school-age children, including the Lidcombe Program (Hofslundsengen et al., 2022; Laiho et al., 2022).

SLPs identified learning needs

The SLPs in our study identified a range of knowledge and skills gaps in their management of paediatric stuttering. These included several aspects related to identifying and measuring stuttering, treatment options, and clinical decision making, particularly when managing complex clients or translating research into typical clinical contexts. Clinical decision making and the need to accommodate and adapt for clients, their families, and the treatment context are consistently reported in the clinical translation literature for SLPs working with CWS (Byrd & Donaher, 2018; Van Eerdenbrugh et al., 2018). These issues do not occur in clinical trials due to defined selection criteria and the homogeneity of the treatment schedule, context, and prescribed treatment protocol. Despite most of the participants only working with a handful of clients each year, the educational needs did not reflect only basic stuttering knowledge or skills (as was found in Tellis et al., 2008). Indeed, the broad range of topics identified reflects the complexity of working with this population in an array of clinical settings. It is encouraging to note the SLP focus on updating themselves with the latest research evidence and evidence-based practice more broadly.

Limitations

While this research has provided an important insight into the practices, perspectives, and educational needs of a large sample of Australian SLPs who work with CWS, there are several limitations of the study. Firstly, the survey used was developed for the purposes of the research and so its validity and reliability have not been established. Secondly, while the survey included a series of open-ended questions designed to probe with more depth, there was no way of further investigating or following up on participant responses. We believe there is potential for future research to utilise a qualitative design that enables a more in-depth analysis. Thirdly, the sample consisted of SLPs who had the motivation and time to engage in a paediatric stuttering professional development workshop. It is therefore likely that our sample is not representative and indeed we may be reporting on a positively biased sample when compared to the broader Australian SLP community. Nevertheless, our sample, survey design, and recruitment procedures were relatively similar to other published investigations of SLPs who work with CWS (Beita-Ell & Boyle, 2020; Boyle et al., 2021; Briley, 2018; Byrd et al., 2020; Coalson et al., 2016; Tellis et al., 2008). Finally, it should be acknowledged that the data were collected over a number of years and research related to the management of children who stutter continues to evolve. As such, the perspectives and experiences of Australian SLPs may have changed. We plan to replicate this study in the future to establish whether this is the case.

Conclusion

This study presents a current investigation of the practices, perspectives, and educational needs of a large sample of Australian SLPs who work with CWS. Our findings corroborate those of previous studies showing that for those SLPs delivering the Lidcombe Program, undertaking formal training equates to higher self-reported success with the program. This is even more important than clinical experience. However, more clinical experience appears to increase an SLP’s confidence in both providing education about stuttering and delivering the Lidcombe Program. SLPs report that there are numerous important factors that can contribute to the success of the program, many of which are attributed to the family and clinical context. Australian SLPs also draw on a range of indirect and direct invention approaches when working with CWS. Additionally, the complexity of working with this population is evidenced by the wide ranging clinical translation issues and educational needs identified. Future research should seek to delve deeper into the experiences and perspectives of this population.

Declaration of interest statement

No potential conflict of interest is reported by the authors.